TogetherCrew: Grow your Farcaster Community,

Public Good Projects Discussion
1

Project name

TogetherCrew

Project Description and why it’s classified as a Public Good

TogetherCrew is a community success tool supporting community builders in their day-to-day tasks with insights and automations. We contribute to a sustainable growth of communities without burning out community builders.

We offer two sets of features: Analytics (View your data) and Actions (take actions based on your data). For example, with our dashboard, community builders quickly know who the active members are and spot trends. To bridge the gap between data and impact, we offer several features that build on top of our analytics: smart announcements, reputation NFT, Hivemind (Q&A-bot).

We kicked off our project in May 2022 with a deep dive into community science. This was funded by Aragon. Since then we have continued to fund our efforts through grants (e.g., Aave, MetaCartel, Optimism, Web3 foundation, SingularityNet). Summer 2023 we launched our app, providing analytics to communities on Discord. Since then we have further developed our backend, increasing the efficacy of data ingestion and computation, and added new metrics to our dashboard. We are currently completing a grant by SingularityNet for developing an LLM bot (HiveMind), and another grant by Arbitrum for developing an off-chain reputation score that is displayed onchain (dynamic NFT). As part of these grants we are integrating data from other community platforms (Discourse, Telegram).

We are building TogetherCrew as an open source project, including our metrics. As PhDs and scientists, we know that the data that feeds our algorithm isn’t ours, but is owned by the community. Our vision is to enable communities to build their own mini-apps powered by their own data. To enable this, we are indexing and structuring the conversations that are happening on various communication platforms. In addition to this, our community engagement metrics adopt a human approach, by considering activity over a longer timespan. For example, members do not need to post every day to be considered active. We know that live happens outside of community platforms.

Main Project Funding Sources

We have received grants and completed work for

Project Funding

We are seeking project-specific funding to support the development of a Farcaster pipeline and features. We have two features in mind:

Know your community

This feature is community-leader centric. The admin of the community can turn on this feature, and decides who has access to it. It allows them to

  1. Receive a quick update on what happened while they were away
  2. Get more familiar with members by knowing their interests

This information can be used to develop better resources for the community (content, events etc), spot collaboration opportunities etc.

How we will do this:* We will summarize conversations using the expertise we already developed thanks to HiveMind. These summaries will be send to the user via direct message. The user will decide the timeframe for these summaries, data source (e.g., channel data, channel follower casts, token holder casts), and when the summary should be received. Only admins of the TogetherCrew app will be able to decide the content of the summary and who should receive it.

FriendRecommender

This feature is community-member centric. It will be available to everyone on Farcaster, but only use data from communities who use TogetherCrew. This is in line with our principle that a user’s data should not be used without their consent.

We will create a frame. This frame provides users with a recommendation with whom to connect with. This recommendation is based on previous casts, and matches users based on their similarity of casts. 6

Roadmap and Milestones

We’re looking for ETH 3 to ETH 5 (USD 18k) in funding to support this development.

Milestone 1: Infrastructure Setup
The pipeline integrates Neynar webhooks, listening to real-time Farcaster events and storing them in a graph database.

Milestone 2: Farcaster Analyzer Pipeline
The pipeline transforms Farcaster data into a data format compatible with the TogetherCrew analyzer.

Milestone 3: Farcaster LLM Pipeline
The pipeline transforms Farcaster data into a data format compatible with the TogetherCrew LLM.

Milestone 4: Community summaries
Design message scheduler and message parameters build bot that sends DC Test bot Deploy bot

Milestone 5: FriendRecommender
Design frame flow + parameters + growth/usage strategy

Team Information

For this project, the following team members will be core:

  • Amin will be responsible for the data analytics and LLM work.
  • Behzad will be responsible for creating the data pipeline: Backend Engineer with over 5 years of expertise spanning IoT, streaming services, and financial social networks.
  • I (Katerina, contact person) and Daniel will be responsible for product development (user experience, research). Cyrille is the tech lead with an eye on making features not overly complex and staying within scope

Our team also consists of a front-end engineer (Nima), marketing lead (Ashish). We are operating permissionless and regularly use bounty hunters for smaller tasks. This has served us well to overcome bottlenecks and onboard new core contributors.

Key team members for this project
Katerina (contact person) - Co-lead TogetherCrew. Ph.D. using social network analysis. Since 2016 she is co-instructing a graduate course on data analytics for HR at Northwestern University. She has also co-organized the Learning in Networks sessions at the International Conference of Social Network Analysis (2018 - 2020), and previously advised a people analytics company on social network metrics. Twitter: twitter.com/katerinabohlec Linkedin: linkedin.com/in/katerinab

Github: katerinabc · GitHub

Danielo (product lead) - Instigator at RnDAO and CoLead at TogetherCrew. Previously, Head of Governance at Aragon, 8 years experience in Organization Design consulting (clients include Google, BCG, Daymler, The UN, and multiple startups), and visiting lecturer at Oxford University. Twitter: x.com LinkedIn: https://www.linkedin.com/in/conductal/

Cyrille Derché - Tech lead at TogetherCrew. Bsc. Computer Science. Ex-Accenture, Co-founder and CTO of SaaS company helping medical device manufacturers deliver product data + documentation to healthcare professionals (handling of sensitive data). Builder of products, processes, and teams.

Social Credibility

We have an active Twitter presence with bi-monthly Twitter Spaces. I’m also active on Farcaster, beginning to build our community on it. On Medium we post articles to help community builders, and discuss new features.

Our app is available via www.togethercrew.com
Our code is open-sourced on GitHub.

Discord contact

Katerina (@kbc)

2 Likes
2

I can’t wait to apply Together Crew to Chones Warpcast channel!!

2 Likes
3

Short-term update

Note: I don’t plan to participate in Epoch 5. We still have to action what we laid out in Epoch 4. This will happen after we have delivered the milestones from other grants.

Total Funds Received (Previous Round): $ / ETH [Amount received in the previous round]
0.1454ETH

Total Funds Requested: $ / ETH [Amount requested from initial application]
3ETH to prioritize this work. We submitted to the GitCoin round (Thrive, and CollabTech). The CollabTech round closed on Tuesday, Thrive today.

Funds Spent: $ / ETH [Detailed breakdown of expenditures]
0ETH

Remaining Balance: $ / ETH [Remaining funds from the previous round]
0.1454ETH

Plans for Next Epoch: Work on Milestone 1

Milestone 1: Infrastructure Setup
The pipeline integrates Neynar webhooks, listening to real-time Farcaster events and storing them in a graph database.

Karma GAP: Error signing in. I’ll update once this is solved

Additional Funding Request: $ / ETH [Total additional funding required]
None at this moment

1 Like
4

Update September 2024

We still don’t have enough funding to execute on the complete idea initially proposed during epoch 4. To overcome this we have participated in Gitcoin round 21, and plan to participate again in round 22.

By the end of September we had the funds to start with the back-end work (farcaster pipeline).

1 Like
5

Update November 2024

  • Total Funds Received from Octant so far: 0.1514 ETH
  • Funds Spent: 0.1514 ETH
  • Detailed Utilization: na

Milestones and Deliverables

  • Milestone 1: Infrastructure setup
    • Description: The pipeline integrates Neynar webhooks, listening to real-time Farcaster events and storing them in a graph database.
    • Immediate Outcome: we have a initial code for pulling in the data and storing it in our database. A couple of changes are still needed.
    • Funding Used: 0.1514
  • Karma GAP: TogetherCrew | Karma GAP

Challenges

  • Previous Challenges:

    • Description: na
    • Updates: na
    • Resolution: na
    • Lessons Learned: na

  • Future challenge: As our team is stretched very thin completing grant work for Singularity Net, Arbitrum and Optimism, we have reviewed our roadmap with an eye on technical complexity and potential for revenue. While integrating Farcaster data is an easier task than Telegram, given their api documentation, we decided to de-prioritize Farcaster and focus on Telegram integration. Our goal is to have the telegram integration ready for ethdenver.

Outputs and Outcomes

Outputs (epoch 5):

  • Output 1: reputation nft - funded via Arbitrum
  • Output 2: script for ingesting farcaster data - funded by Octant

Short-term Outcomes:

  • Outcome 1: We’ve gained 10 new users in the past month. 1 user switched to a premium subscription.
  • Outcome 2: We’ve developed a tighter feedback loop with new users, reaching out to them more regularly. We’ve also switched to a high-touch customized outreach message to communities which has helped us understand the user needs better.

Long-term Outcomes:

  • Outcome 1: User engagement: We have 4 communities who regularly check their metrics (+2 since the summer). We have also grown our social media presence and have a small telegram community where we talk more about community management practices, data, and our product.
  • Outcome 2: Innovation: With the release of our AI bot we have achieved a major milestone in our technical capabilities. The next step is to combine the AI technology with our analytics to provide deeper insights.

Plans for Potential Funds Raised in the Next Octant Allocation Window

  • Goal 1: Telegram
    • Description: Integrate Telegram into our community management app. This requires ingesting the data, structuring it according to our schema, and adding a new login method for Telegram.
    • Expected Outcome: Communication in Telegram is high-paced. Many communities have moved from Discord to Telegram. Statistics solution for Telegram are less developed than for Discord. We’re expecting an increase in growth of active users.
    • Funding Needed: 10k$
  • Goal 2: Farcaster
    • Description: This is a stretch goal and only possible if we get the necessary funding and are able to attract a high quality bounty hunter to support us in the work.
    • Expected Outcome: Given how active the farcaster community is, we’re expecting the feature to help us build an active community, get feedback on what is useful for community members.
    • Funding Needed: 7.5k $

Other Funding

  • Grant Funding Received Since the Last Epoch: Gitcoin round 21 (CollabTech round) and round 22 (tools)
  • Other Non-Grant Funding Sources: 2 premium users.

Future Plans for Financial Sustainability

We are continuing the route of grant funding. We are looking to change the balance from dominantly grant funding to revenue from clients. We are exploring partnerships with agencies.